Until 2008, calibration was performed using a traditional thermopile pyranometer as the reference instrument, as it represented the only widely available and reliable solution at that time. During this period, our primary photovoltaic (PV) sensors were calibrated by comparison with a thermopile pyranometer, specifically a Kipp & Zonen Secondary Standard, recognized for its high measurement accuracy. However, extended comparative field testing revealed limitations associated with this approach. In particular, calibrations carried out during very hot days or around solar noon were found to be affected by systematic discrepancies. Under these conditions, the infrared (IR) component of solar radiation increases significantly; thermopile pyranometers respond to this IR contribution, whereas PV reference cells do not, due to their photovoltaic conversion mechanism. As a result, the use of thermopile instruments as calibration references for PV sensors can introduce systematic errors in the derived irradiance values.
In parallel with these empirical observations, the publication and subsequent evolution of the IEC 60904 series provided a formal framework for the design, characterization, and calibration of photovoltaic irradiance sensors. In particular, IEC 60904-2 defines traceable calibration procedures for PV reference devices, explicitly accounting for their spectral responsivity, temperature dependence, and operating conditions. The introduction of these standards clarified the appropriate calibration methodology for PV-based irradiance sensors and addressed the limitations inherent in thermopile-based reference calibrations for photovoltaic applications.
Today, these approximations are no longer acceptable. The photovoltaic sector has evolved and matured. When assessing the performance of a PV plant, we can no longer ignore the fact that specific standards now exist for the traceability and calibration of PV reference cells standards that did not exist or were not widely adopted years ago.
Moreover, under certain conditions, relating PV performance to a thermopile sensor can be less accurate than using a calibrated PV reference cell. The market now demands higher precision in estimating and controlling the production of plants ranging from hundreds to thousands of kilowatts. A typical error caused by the IR fraction undetectable by PV cells often around 4% is no longer tolerable.
Calibration of Photovoltaic Pyranometers
The calibration procedure follows the requirements of IEC 60904-2, Edition 4.0 (2023), the latest and most up-to-date version of the international standard for the calibration of photovoltaic (PV) reference devices.
This edition introduces important updates, including extended procedures for maximum-power calibration, revised requirements for spectral responsivity and temperature coefficient measurements, improved specifications for built-in shunt resistors, and explicit rules for calibration traceability.
Calibration is performed by comparison, as defined in IEC 60904-2.
The reference device is a primary reference sample calibrated at the ISH institute, accredited by DAkkS, using an “extended” calibration procedure that provides a total uncertainty of 1.0%.
The calibration of our thermopile pyranometers is performed in accordance with ISO 9847.
Calibration is carried out by side by side comparison with a reference instrument. Our primary reference is a secondary standard pyranometer from a well-known manufacturer, calibrated by the factory and traceable to the primary reference pyranometer of the World Radiometric Centre (WRC).
The calibration process is performed under natural sunlight, either outdoors in suitable weather conditions or through controlled exposure in the laboratory. The output of the primary pyranometer (e.g., SR20) is recorded by a data logger with 14‑bit resolution, which also applies temperature compensation.The pyranometers under calibration are connected to the remaining logger channels. Instead of using only two calibration points, we acquire at least 315 measurement pairs. The calibration factor is then determined using the mathematical procedure described in ISO 9847. Finally, the irradiance readings are recalculated using the new calibration factor to obtain the updated instrument sensitivity.
The adopted calibration approach is fully consistent with the principles and recommendations of ISO 9847 for outdoor pyranometer calibration. In particular, the use of a higher-accuracy reference instrument with documented traceability, side-by-side mounting geometry, and nearly simultaneous data acquisition ensures compliance with the standard’s requirements for minimizing systematic errors. Furthermore, calibration measurements are performed under stable environmental conditions and over sufficiently long time intervals, as prescribed by ISO 9847, to account for the influence of atmospheric variability, angle of incidence, and temperature effects on measurement uncertainty. The resulting sensitivity values are therefore valid for the calibration conditions and provide a reliable basis for subsequent irradiance measurements in operational use.
